Anti-Cryptosporidium efficacy of BKI-1708, an inhibitor of Cryptosporidium calcium-dependent protein kinase 1

Abstract

Background: Diarrheal pathogens, such as Cryptosporidium, impose a heavy burden of disease in resource-limited regions. Cryptosporidiosis often causes chronic infection in immunocompromised people and gastrointestinal injury in malnourished children, leading to wasting, stunting, and cognitive impairment. Current treatment for cryptosporidiosis fails in these vulnerable populations, highlighting the need for new medicines. Here we describe the anti-Cryptosporidium efficacy, pharmacokinetics, and safety of a bumped kinase inhibitor BKI-1708. BKI-1708 inhibits the essential molecular target, calcium-dependent protein kinase 1 (CDPK1), which is highly expressed in the major proliferative stages of the parasite life cycle.

Methods and findings: Efficacy was demonstrated in the Cryptosporidium parvum IFNγ-KO mouse infection and calf diarrhea models. Dose response in the mouse model demonstrated oral doses as low as 15 mg/kg administered daily for 3 days completely suppressed oocyst shedding. Metabolite profiling in pre-clinical species and human hepatocytes identified an active metabolite, M2, which retains sub-micromolar activity against C. parvum. Pharmacokinetic analysis of BKI-1708 and M2 in mice demonstrates good systemic exposure, important for treating biliary and upper respiratory infections in some cases of cryptosporidiosis. In mice, M2 reaches 7-fold and >3-fold higher levels over BKI-1708 in plasma and the gastrointestinal tract, respectively. Oral administration of M2 completely suppressed oocyst shedding in the mouse model at doses as low as 8 mg/kg for 3 days. Wide safety margins are demonstrated in mice, rats, and dogs.

Conclusions: BKI-1708 has characteristics of a safe and effective drug for treating Cryptosporidium infections in animal models and shows promise for use in humans. Moreover, BKI-1708 and M2 formed in vivo, offer an attractive prospect of a dually active preclinical candidate for the treatment of cryptosporidiosis.

Fig 1. Targeting the Cryptosporidium life cycle with BKIs.

(A) Life cycle of the Cryptosporidium parasite. Upon infection, the parasite undergoes 3 rounds of merogeny (green) followed by gametocytogenesis (blue) and oocyst formation (purple). CDPK1, essential for parasite survival, is expressed during rapid asexual proliferation (green). (B) BKI-1708 structure and previously reported C. parvum CDPK1 IC₅₀ and nanoluciferase-expressing (NLuc) C. parvum EC₅₀ values [29].

Fig 2. Visualizing the effects of BKI-1708 treatment of C. parvum infected monolayers.

(A) SEM of untreated HCT-8 monolayers infected with C. parvum. Right panel shows egressed merozoites. Bar: left, 2.5 μm; right, 1.4 μm. (B) TEM of vehicle-treated monolayers show various stages of merogony. pvm: parasitophorous vacuolar membrane. Bar: left, 0.8 μm; right, 0.8 μm. (C) SEM of infected monolayers after 24 h treatment with BKI-1708. Bar: left, 2.5 μm; right, 1.6 μm. (D) TEM of BKI-1708 treated monolayers. Bar: left, 1.0 μm; right, 1.0 μm.

Fig 3. BKI-1708 efficacy in a mouse model of Cryptosporidium infection.

(A) Experimental design of the NLuc C. parvum IFNγ-KO mouse model. Mice (n = 3) are orally challenged with 10⁴ NLuc-expressing C. parvum oocysts on day 0 and treatment commenced on day 6. Fecal samples are collected regularly until day 20 p.i. for NLuc detection of oocyst shedding. (B) Results of 3-day BKI-1708 treatment. The dotted line at 300 RLU denotes the limit of detection (LoD) for this assay. The untreated control group (dark gray closed circle, solid line) in this experiment were euthanized on day 13 p.i. due to excessive weight loss and moribundity. The control (light gray open circle, dashed line) used for all mouse efficacy experiments was 100 mg/kg BKI-1369 administered for 5 days, a former late-lead candidate for cryptosporidiosis [27]. (C) Results of BKI-1708 administered at 15 mg/kg for 1, 2, and 3 days. (D) Results of single dose BKI-1708 treatments.

Fig 4. Metabolite identification in primary hepatocytes.

(A) Probable biotransformation products of BKI-1708 determined via incubation in cross-species hepatocytes. Dehydrogenation and O-glucuronidation appear to be the main metabolism pathways. The dehydrogenation product, M2, undergoes isomerization to form a more stable, cyclized hemiaminal. (B) Ion chromatograms showing relative abundance of biotransformation products from cross-species hepatocytes. Top to bottom, results are from mouse, rat, dog, monkey, and human hepatocyte incubations. M2 appeared to have the highest abundance in mouse and human incubations and was synthesized for further characterization.

Fig 5. In vivo exposure of BKI-1708 in mice.

(A) Plasma exposure of BKI-1708 and M2 after 30 mg/kg PO administration of BKI-1708 in Balb/c mice (n = 3 per time point). (B) Similar to plasma, exposures from homogenized brain tissue show M2 accumulating to levels >3-fold higher than the parent molecule. (C-F) Exposure in GI sections reveal BKI-1708 levels are highest within the first 2 h post administration while M2 levels build up to 6 h and remain at or above 1 μM. (G) timsTOF MALDI-MSI showing strong BKI-1708 presence (green) in mouse small intestine and colon sections 1 h post BKI-1708 administration (n = 4). Signals were overlayed on hematoxylin and eosin (H&E) stained tissue slides of post-imaged sections. (H) QQQ MALDI-MSI showing BKI-1708 and M2 presence (yellow) in mouse small intestine and colon sections 24 h post BKI-1708 administration (n = 4). Phosphatidylcholine headgroup PC(34:1) (blue) was used for detection of tissue-containing regions. Signals were overlayed on H&E-stained tissue slide sections. BKI-1708 and M2 signals show accumulation in the mucosal layer at the luminal surface. (I) QQQ MALDI-MSI showing BKI-1708 and M2 presence (yellow) in the colon at 1 h versus 24 h post BKI-1708 administration (n = 4). At 24 h, the signal remains in the lumen, but is mostly absent from the intestinal wall.

Fig 6. Efficacy of M2 in a mouse model of cryptosporidiosis.

Mice (n = 3) are orally challenged with 10⁴ NLuc-expressing C. parvum oocysts on day 0 and treatment commenced on day 6. Fecal samples are collected regularly until day 20 p.i. for NLuc detection of oocyst shedding. M2 was administered PO for 3 days resulting in significant reduction in oocyst shedding for all doses compared to untreated controls. The control (light gray open circle, dashed line) used was 100 mg/kg BKI-1369 administered for 5 days, a former late-lead candidate for cryptosporidiosis [27].

Fig 7. Efficacy of BKI-1708 in a calf model of cryptosporidiosis.

(A) Experimental design of the newborn calf clinical model of cryptosporidiosis. Newborn calves (n = 3 treated; n = 16 untreated) were orally challenged with 5x10⁷ C. parvum oocysts on day 0 and twice daily 5 mg/kg BKI-1708 treatment commenced on day 2 for 5 days. (B) Mean percent change in weight from birth to the end of the experimental period, on day 10. (C) Mean daily urine output. (D) Daily fecal volumes. (E) Daily fecal consistency scores. Consistency scores range from 1 for normal-formed stool to 4 for severely diarrheic stool. (F) Daily clinical evaluation scores. Clinical evaluation scores are a summative assessment of clinical health parameters (willingness to rise, stance, appetite, attitude, and hydration status), where lower scores reflect normal/healthy calves [67]. (G) Mean daily oocyst counts. Fecal samples were collected daily until day 10 p.i. for qPCR detection of oocyst shedding (n = 3 technical replicates).